In this episode, Brian McKenna, enterprise applications editor at Computer Weekly, and Paul Roehrig, chief strategy and marketing officer at software engineering services firm Ascendion, discuss the likely future political economy of artificial intelligence (AI), deliberately avoiding the twin dangers of catastrophism and boosterism – the Scylla and Charybdis of contemporary AI discourse.
In the first of a three-part series, they explore how AI could transform the economy, business models and employment. Rather than viewing AI as either catastrophe or a utopia, a balanced, evidence-based approach built around a “beginner’s mind” is more useful, argues Roehrig.
Download this podcast
Brian begins the discussion by asking what the shape of a plausible counter narrative to the negativity that surrounds AI could be, especially among the young, where people believe it will destroy jobs, creativity and even civilisation. On the other hand, the AI enthusiasts among the tech elites of Silicon Valley and elsewhere expect it to create a perfect future. Paul argues that neither extreme is helpful. Previous technological revolutions generated similar uncertainty, he says, and society is still learning how to understand AI. The most realistic approach is therefore to examine its opportunities and risks objectively rather than becoming either excessively optimistic or pessimistic.
Paul puts forward the idea that AI will change how we do things rather than fundamentally change what people want. Humans value social interaction, entertainment, shopping, better healthcare and safer financial systems. AI’s significance lies in changing the processes through which these objectives are achieved, he says. It can alter how organisations operate, how work is performed and how new business models are created.
Paul further argues that AI can reduce the technological and organisational “debt” that has accumulated in businesses over decades. Old systems and processes often create unnecessary friction and are expensive to change. Applying AI to workflows can reduce this friction, freeing capital, time and human creativity for innovation.
He gives an example from his own firm’s practice of a financial services company with hundreds of thousands of lines of legacy code dating from the 1980s. Using AI agents, the company was reportedly able to modernise the platform in about half the time and for roughly one-third of the cost.
The discussion also considers whether AI should autonomously redesign business processes. Paul rejects the idea that humans should surrender intellectual control to machines. Instead, he advocates human-AI collaboration.
Contact centres
The two discuss contact centres as an example of how human-AI collaboration could work, building on some work Brian did as an analyst at Informa …’s Enterprise Strategy Group. AI could handle repetitive, data-intensive tasks while human employees focus on complex problems, empathy, customer relationships and sales. This could change the traditional economic model of contact centres.
Historically, businesses have treated them mainly as cost centres, seeking to minimise employees and shorten calls. AI could instead help transform them into growth engines, where human workers use AI to improve customer experiences, cross-sell products and generate revenue. Some jobs will undoubtedly be at risk, particularly those involving highly routine work, but AI does not automatically mean fewer employees in every area.
In discussion, Paul compares AI with electricity rather than simply the internet. The internet transformed existing products and services, whereas electricity changed the fundamental means of production. Similarly, AI could become a new “power source” for intellectual and cognitive work, enabling organisations to redesign operations and develop new economic models.
The discussion concludes that society should neither blindly embrace nor reject AI. Instead, humans should approach it with curiosity and a beginner’s mind while maintaining responsibility and control.
AI has significant risks, and some things will inevitably go wrong, but it also offers opportunities to improve productivity, innovation and economic outcomes. Its ultimate impact will depend not simply on the technology itself, but on how intelligently humans choose to use it.
S 004